03-20-2019 06:07 AM
If you have concern on performance, open a different thread may get better information and help especially if others have the same problem.
Some clue like enable performance mode, put a clean image on it, try different video driver...etc.
03-20-2019 06:54 AM
the point was the performance with bios 1.19 is LOWER than with 1.18
and the machine is clean - and performance is set to maximum.
03-20-2019 07:03 AM
Hmm...for now, you have the choice to roll back to 1.18 if you believe in what Cinbench tells you.
We occasionally fall into the situation that some benchmark software tells something worse but that usually not able to proved as a real problem.
03-20-2019 10:05 AM
I appreciate your engagement in the community, but I must say I find your reply there a bit strange.
Computing performance is quantitative; it is measurable and graphable with software such as HWiNFO, 3dmark, and cinebench.
Just as fans can be measured in decibels and RPMs, performance of the computer is objectively measured with cinebench multi 64 - often in loops to simulate extended multi-threaded work that most of do frequently. The score of the CPU is very important here, because it suggests that Lenovo's engineers have lowered the maximum performance the machine is capable of with this BIOS patch. Consumers have a right to be upset by that (and I will go on record as one of them).
It's fine to adjust fan curves, but fans should always max out to reduce temperatures rather than lower performance. If you want to make the machines quiet for some, then it's best to tune a quiet fan option for lower performance and higher temperatures in the Lenovo Vantage software. Dell has quite a good tool for this on their XPS PCs, actually. Leave the "maximum performance" at maximum performance, please.
03-20-2019 10:25 AM - edited 03-20-2019 10:30 AM
Cinebench CPU is a stactic CPU test. If CPU had plenty of cooling - each result should be able to replicate within a few "points" - just like Intel XTU should.
Cinebench and XTU is usually not impacted by crap installed on the machine like SysMark would be or other equivalent "overall system benchmarking tools"
The only thing that can draw drown performance is either bad cooling or other resources taking CPU cycles away.
Btw. I am a former journalist & editor of a big brand PC publication + about 30+ years PC hardware experince. (Incl. hw design)
@Henrydong And I do find your comments a bit uneducated.
03-20-2019 02:53 PM
I just ran mine. 965 Marks. Every kind of throttling going on during the test. I wear hearing aids but I never hear the fan running. The only way I do is if I hold the bottom of the laptop up to my ears.
I ran the benchmark and adopted the suggestions from HWBOT. It improved to 983 Marks, but one of the changes was to crank up Turbo Boost Short Power Max to 60W. Where can I see the fan speed? Like I said above, I don't hear anywhere near as much noise as others are talking about. I wonder if my fan control isn't working right with 1.17.
03-20-2019 04:51 PM
Out of the box, BIOS flash to 1.19 and then a flash of the OS for a clean install of Windows 10 Pro. Average temperatures are sitting at around 38-48'C from the CPU package sensor with it peaking at 70'C during compiles running all cores. I've also run a few tests using Tensorflow and the temps reached 99'C briefly. Battery life is decent so far, roughly 6hrs of battery with about 25% left, but my workloads mostly involve Visual Studio with remote debugging sessions in C/C++ and Python for the ML stuff.
Knock on wood, pretty happy with these Thinkpad machines. Very different build quality from the consumer lineup.
03-20-2019 05:11 PM - edited 03-21-2019 11:30 AM
The 1.19 update came to my P1 Xeon today via Vantage.
I installed it - but no big change. XTU score is still low - fans still spinning high - and Vantage still reports Bios 1.18 and no new updates.
During the update something flashed on screen (built in) in very small letters and then shut down. After 5 min I turned on again and it proceeded to update some microcode and reboot.
I also installed it via Vantage today, and noticed the same issue where Vantage still reported version 1.18. After I noticed that, I rebooted into the bios to check the version there. The bios correctly reported version 1.19. Then I rebooted back into Windows, and now Vantage correctly reports version 1.19 also. So I think Vantage was just temporarily confused.
Even when Vantage was reportign the wrong version, I could tell that the fan behavior had changed, so I think it was just a status display error.
03-20-2019 06:28 PM - edited 03-21-2019 11:52 AM
I am pleased with bios version 1.19 on my P1. Fan noise is definitely reduced, and I don't see any negative effects on temperature or performance. The lowest fan speed is now about 2200 rpm, which is very quiet. Good work Lenovo!
Edit to add:
I my office where there is some background noise, and I'm working on external keyboard/monitor, the 2200 rpm fan speed is inaudible. If I'm working directly on the laptop in a quiet room, it is still audible. It also seems like the fans turn on at low speed more frequently with bios 1.19 than with earlier versions, even during low CPU load, so in that situation the new bios may produce more fan noise than older versions.
Bios 1.18 Cinebench R15 - 965 (CPU Multi) - reasonably consistant.
Bios 1.19 Cinebench R15 - 905-955 (CPU Multi) - quite random clean booted machine. But a lot more random than 1.18 bios
Reviews of the P1 -
Notebook Check : Cinebench R15 - 1118 (CPU) -https://www.notebookcheck.net/Lenovo-ThinkPad-P1-Xeon-E-2176M-Quadro-P2000-Max-Q-Workstation-Review....
PC Magazine : Cinebench R15 -1149 (CPU) - https://uk.pcmag.com/laptops/118584/lenovo-thinkpad-p1
The point was the performance with bios 1.19 is LOWER than with 1.18
My experience is different. Bios 1.19 doesn't appear to cause any performance degradation for me. I did multiple runs of Cinebench R15 on bios 1.18 just before the update, and on bios 1.19 just after the update, and I don't see any signficant difference.
I tested after doing a reboot, no other programs running, no external devices connected except a mouse, and I'd wait for the fan to ramp down between runs. Even so, Cinebench didn't give me consistent results with either bios version. With both versions I saw a wide range of results from about 920-1020, usually 950-1000.
Some important variables to keep in mind when comparing benchmark numbers:
1) Machine config: I have the i7-8850, P2000, 16GB single channel RAM, 4k display.
2) Undervolting: I tested without any undervolting. A modest undervolt can increase the score by about 10%.
3) Hyper-V: I had Hyper-V enabled. Currently if Hyper-V is enabled in Windows, that causes the CPU to throttle at 80C instead of 100C, which reduces the score by about 10%. Most reviewers probably had Hyper-V disabled, since that is the default, so their scores will be higher than yours if you have Hyper-V enabled.
For more discussion on the Hyper-V issue, see this thread:
4) Windows patches for issues like Spectre and Meltdown can reduce system performance, making older benchmark scores look better than current scores.
03-22-2019 11:24 AM
My X1E was also bricked during the BIOS update so this is not a rare or an isolated issue as Lenovo is advising. I ended up returning it because it scares me if it happens after my warranty expires.